30 On the Measure of Temperature. 
De Luc and Crawfwd include too limited a portion of the thermometric scale, to 
allow of o r deducing from them any general results. And this is a defect 
common to all the inquiries into the theory of heat yet made. M e may, m fact, 
easily conceive, that phenomena subject to very different laws, may yet, within a 
certain interval of temperature, appear to have the same progression. For, if we 
content ourselves with observations, between tl»ose limits where the divergence of 
the laws is scarcely sensible, we shall be naturally led to attribute such small 
differences to error of observation ; and we are thus left without data for investi- 
gating their real cause. We shall have frequent occasion, in the course of this 
memoir, to acknowledge the truth of this conclusion. ... , 
Mr. Dalton has considered this question from a more elevated point oi view, and 
has laboured to establish general laws, applicable to the measurement of every 
temperature. These laws, we acknowledge, form an imposing whole by their regu- 
larity and their simplicity. Unfortunately, this aide philosopher has been too 
hasty in generalizing — conceptions' highly ingenious, it is true, but which rested 
upon very uncertain data. Accordingly, there is scarcely one of his positions, 
that is not contradicted by the results oi' the experiments of which we are about 
to give an .account. 
These experiments have had for object, to discover the lairs of cooling, to 
which bod es are subje t when immersed in an elastic fluid* of a certain nature, 
density, and temperature. Before commencing this inquiry* however, it was 
necessary to supply the perfect blank which exists, and thus render more precise, 
our notions on the’ measure of the higher temperatures. We therefore commenced 
our labours with an examination of this question, secondary, it is true but still 
itself possessing great iriterest ; and we propose to enter in the first instance on 
the details connected with this subject. 
This p iper then, will be divided into two parts perfectly distinct : in the one, 
we propose to investigate the measure of temperature ; in the other, the laws of 
cooling. 
PART I. 
On the Measure of Temperature. 
If we could find a substance, the expansion of which were regulated by a law 
so regular and simple, that the addition of successive quantities of heat should he 
accompanied bv equal increments of volume; such a substance would possess all 
the qualities which philosophers have been accustomed to look for in a perfect 
thermometer. 
Such an instrument might, however, he found not to answer all the expectations 
at first excited. If it should so happen, for instance, ihat the specific heat of 
every other b dv when referred to its scale, proved variable, and unequally variable 
ill each ; it is evidetit that no conclusion could he drawn a priori, from the indica- 
tions of this instrument, as to the quantities of heat gained or lost in a given change 
of temperature. 5 
We see then, that our first step in this inquiry, is to ascertain whether the 
capacities of a number of bodies referred to the same scale, vary in the same man 
ner. and whether the expansions of substances of the most opposite natures are 
regulated by one and the same law. This latter comparison, witli which we shall 
begin our details, being susceptible of greater precision titan the first we have 
entered into it much more fully ; and we do not .believe, we have neglected anv of 
the precautions, necessary to ensure correctness in the results. 6 J 
$ 1 — On the Expansion of the Gases. 
of^ns- m the — 
which the different valuations shall be referred, is evidently to'be^etmSned on 
considerations purely arlutrary. Superior facility of construction and greTter con 
v entente ... observing, were our inducements for employing the mercuriT hermo 
meter m almost ail our experiments. 3 & *ucicunai tneimo- 
o/noLX’ta^^ be conveniently confined to liquids 
of nomenclature in Science, is Precision 
